Role of Artificial Intelligence in Object Detection: A Review

Over the last 20 years, the field of object detection has seen a remarkable progress been made. In its 20-year history, this literature review not only goes through some landmark detectors (e.g., VJ detector, HOG detector, DPM, Faster-RCNN, YOLO, SSD, and others), key technologies, speed up methods, detection applications, datasets, and metrics, but also the 28 difficulties currently faced by the community are being discussed and further enhancement and continuation of these detectors are also discussed. We had shallow trainable structures and Handcrafted features which are the fundamentals of conventional object identification systems. With the spontaneous advancement of deep learning technology, additional effective tools are now competent of learning semantic, high-level, and deeper characteristics. In this study, deep learning-based object detection system has been reviewed. We also go through a few specific tasks, such as prominent item detection, face identification, and pedestrian detection.